Having spent the summer hanging out of a window my PM sensor now has a permanent outdoor home.

The SDS011 is an awkward size. It is just too large to fit inside what turns out to be a standard UK drainpipe. Most B&Q stores will not be able to help you. Online stores selling 110mm pipes will, for a price. I discovered that Wickes also stock them - good news if you are a bricks and mortar type.

Final design

So, after a certain amount of trial and error I have now happily arrived at a final design. I have very closely followed the Luftdaten.de design. The main modifications are:

1) Hydrophobia

I have housed my nodeMCU inside its own waterproof container inside the u-bend pipe. The waterproofing is breached by the hole necessary to bring in the power wires and wires to sensors. I have attempted to keep the hole as small as possible and to protect it against direct water ingress. The nodeMCU generates a small amount of heat during normal operation, which should keep the inside of the waterproof container it occupies a few degrees warmer than the ambient temperature, thereby driving away any moisture that would otherwise gather on the device causing corrosion.

2) Arachnophobia

The end of the u-bend pipe have been covered with mesh (a plastic mosquito mesh equivalent). This is in an effort to keep the number of spiders taking up residence inside the u-bend to a minimum.

3) Acrophobia

How to avoid falling and smashing into a million pieces? Fixture philosophy: KISS. Cable ties FTW.

Location

The location of the installation was governed by 1) security, 2) ease of access for maintenance, 3) proximity to a power supply, 4) shade from direct sunlight at every part of the day.

1) Security

The location needed to not be easily accessible from the road. I would prefer no one walked off with it.

It also needed to be protected against interference from low flying UFOs. Footballs and the like are startlingly effective in reaching all manner of apparently unlikely locations, so if you have children this is an important consideration.

2) Maintenance

This is fairly obvious. I would love to think that my design is flawless and will work perfectly perpetually, but the realist in me knows this is not true. I want to be able to access the device relatively easily.

3) Proximity to power

I do not have any external power sockets. Instead I have brought power to a point inside the house which is close to the exterior installation point. The mains-to-5V DC adapter is plugged in there. Power is routed externally through the eaves on a short cable run.

4) Shade

I want the temperature readings to be as accurate as possible, and I do not want the SDS011 to overheat in the summer, so I have used a section of wall which is shaded throughout the day regardless of the time of year. Harder in some cases than it might sound.

In my case the qualifying location is a haven for spiders and so I have taken the precaution of covering the ends of the pipes with a plastic mesh, secured by hi-tech rubber bands. The mesh still allows free flow of air, and the holes, although intentionally small on the scale of spiders, are vast on the scale of particulate matter pollution, so I expect the impact on results to be negligible.

Conclusion

It has taken a while to get things set up, and I have learned a number of things the hard way. I am glad to have the sensors installed in a low maintenance configuration now, and I look forward to gathering in some interesting data.